Angle of arrival detection using compressive sensing

The conventional method for determining target angle of arrival with an array of sensors is to digitize the output of each sensor at the Nyquist sampling rate for the system bandwidth and use digital signal processing algorithms such as maximum likelihood estimation or multiple signals classification (MUSIC). Here we show that if the targets sparsely populate the angle/frequency domain, the angles and frequencies can be obtained from a much smaller number of measurements by randomly summing sensor outputs and by randomly sampling in time through the use of a novel application of recently developed compressive sensing algorithms.